Introducing HYDRA2 – a Multispectral Data Analysis Toolkit for
Suomi NPP and EOS
Tom Rink, W. Paul Menzel, Liam Gumley, and Kathy Strabala
Space Science and Engineering Center
Madison, Wisconsin, 53706 USA
Abstract
A freeware based multispectral data analysis toolkit for satellite data has been developed
to assist research and development of remote sensing applications as well as education
and training of remote sensing scientists; it is called HYDRA2 - HYper-spectral data
viewer for Development of Research Applications version2. HYDRA2 provides a fast
and flexible interface that allows users to explore and visualize relationships between
sensor measurements (displayed as brightness temperatures for infrared and reflectances
for visible and near infrared) and wavelength (or wavenumber) using spectra diagrams,
cross sections, scatter plots, multi-band combinations, and color enhancements on a pixel
by pixel basis with full access to the underlying metadata of location and time.
HYDRA2 enables interrogation of multispectral (and hyperspectral) fields of data so that
(a) pixel location and spectral measurement values can be easily displayed; (b) spectral
bands can be combined in linear functions and the resulting images displayed; (c) false
color images can be constructed from multiple spectral band combinations; (d) scatter
plots of spectral band combinations can be viewed; (e) pixels in images can be found in
scatter plots and vice versa; (f) transects of measurements can be displayed, and (g)
soundings of temperature and moisture as well as spectra from selected pixels can be
compared.
HYDRA2 is a toolkit that can be used with direct broadcast and archived data from
Suomi NPP and the EOS Aqua and Terra platforms.
1.
Design Philosophy of HYDRA
HYDRA2 (Hyper-spectral data viewer for Development of Research Applications
version 2) grew out of the necessity to provide research scientists and educators with
freely available software that could display and investigate remotely sensed multispectral
and hyperspectral data. HYDRA2 is a fourth generation application produced at the
University of Wisconsin’s Space Science and Engineering Center (SSEC). The first three
were based on commercial packages IDL, Matlab, and Java; they served as a testbed to
refine the operations and functionality requested by the users.
When considering the design of HYDRA2, the most important requirements were that the
software must be (a) freely available to the global community, (b) computer platform
independent, and (c) extendable. Computer platform independence has been elusive in so
far as HDF libraries still are required to view the data. To this end the Java-NetCDF
library from Unidata, which provides a platform independent, multi-array access API to a
wide range of file storage formats, is employed to provide access to HDF4/5, used for
NASA EOS and SuomiNPP. The VisAD (Visualization for Algorithm Development)
Java library was selected as the basis for HYDRA2; Hibbard et al (2002) has more
information about VisAD. This library allows applications to be written in either Java or
Jython, the Java implementation of the popular Python scripting language; more details
are available in Bill (2002) and Lutz et al (1996). The core of the VisAD library is a
unified Data Model, which allows for the representation of literally any numeric data in a
consistent manner. This is very important when integrating data from a variety of
sources that may have different sampling topologies in space or time. The Data Model
also has built-in metadata for parameters such as units and error estimates. Because these
metadata are an integral part of the data, they can be easily used to verify the validity of
computations (the software detects invalid combinations, such as adding a temperature to
a pressure) and to estimate the reliability of computations.
HYDRA2 is a standalone client application organized into a simple hierarchy of Java
classes leveraging the VisAD library, the aforementioned Java-NetCDF library and a
higher level, VisAD based library from Unidata for handling map projections and geo
navigated displays. HYDRA2’s design layout separates classes responsible for generating
displays, interactive data interrogation, custom analysis tools and provides a
programmatically extendable interface for adapting physical data to VisAD’s abstract
data model.
The programmatic interface to the data files has been designed so that emerging new file
types or formats will be easy to support without changing the core of the applications.
The key was to abstract the reading of the files so that the application makes the same
function or method calls regardless of the file being read. This same philosophy is used
in packages such as OpenDAP and ESML, which have the notion of a plug-in.
2.
Getting and Setting Up HYDRA2
The latest stable version of HYDRA2 can be obtaining from the following ftp site:
ftp://ftp.ssec.wisc.edu/pub/CSPP/beta/hydra2/
For Windows Operating Systems (OS) (XP, VISTA 7 and 8) download the ‘exe’ file for
latest version using binary ftp mode, and then run it by double-clicking the file icon. The
installer will ask where you wish to install the program. Accept the default, C:\Program
Files\HYDRA2. Start the program by selecting Start | All Programs | HYDRA2 |
runHYDRA. A window named “HYDRA” will appear (Figure A1); it will indicate the
version you are using. A window named “runHYDRA” will also appear (this window
may be minimized, but do not close it).
For Mac OSX use the ‘dmg’ installer and simply double-click to install into the
/Applications folder. To start HYDRA2, double-click the icon in that folder. The icon can
also be dragged to the applications docking station after which a single click of the icon
will start the application.
For the Linux OS (64bit), the ‘sh’ installer will extract into the directory of choice, or use
the ‘tar.gz’ and install by entering “gzip -cd hydra_v1.5_linux.tar.gz| tar xvf – ” at a
command prompt. This will create a sub-directory named “hydra” into which everything
required to run will be installed. To run HYDRA, first change directory into the 'hydra'
directory, and enter “./runHydra” at the command prompt.
Acknowledgements
This paper relied on the active participation in beta testing the HYDRA2 freeware by
many students and international colleagues. We thank them for their enthusiasm and
useful feedback. Funding from the NASA IMAPP (International MODIS and AIRS
Processing Package) and the NOAA CSPP (Community Satellite Processing Package)
programs is gratefully acknowledged.
References
Aumann, H.H., M. Chahine, C. Gautier, M. Goldberg, E. Kalnay, L. McMillin, H.
Revercomb, P. Rosenkranz, W.L. Smith, D. Staelin, L. Strow and J. Susskind,
2003: AIRS/AMSU/HSB on the Aqua Mission: Design, Science Objectives, Data
Products, and Processing Systems. IEEE Transactions on Geoscience and Remote
Sensing, 41, 253-264.
Barnes, W. L., T. S. Pagano, and V. V. Salomonson, 1998: Prelaunch Characteristics of
MODIS on EOS-AM1. IEEE Transactions on Geoscience and Remote Sensing,
36, 1088-1100.
Bill, R. W., 2002: Jython for Java Programmers. New Riders Publishing. ISBN 0-73571111-9. Pp. 465.
eoPortal https://directory.eoportal.org/web/eoportal/satellite-missions/a/aqua
eoPortal https://directory.eoportal.org/web/eoportal/satellite-missions/m/metop
eoPortal https://directory.eoportal.org/web/eoportal/satellite-missions/s/suomi-npp
eoPortal https://directory.eoportal.org/web/eoportal/satellite-missions/t/terra
Hibbard, W., C. Rueden, S. Emmerson, T. Rink, D. Glowacki, T. Whittaker, D. Murray,
D. Fulker, and J. Anderson, 2002: Java Distributed Objects for Numerical
Visualization in VisAD. Communications of the ACM, 45, 160-170.
Klaes, K. D., and Coauthors, 2007: An introduction to the EUMETSAT polar system.
Bull. Amer. Meteor. Soc., 88, 1085–1096.
Lutz, M., G. Rossum, L. Lewin, and F. Willison, 1996: Programming Python. USENIX.
Mark Lutz Programming Python O'Reilly & Associates, 1996. ISBN 1-56592197-6. Pp. 880.
Murphy, R. P., P. E. Ardanuy, F. DeLuccia, J. E. Clement, and C. Schueler, 2006: The
visible infrared imaging radiometer suite. Earth Science Satellite Remote
Sensing, vol. 1, 199-223. Springer Verlag, 2006.